Filmscanners mailing list archive (firstname.lastname@example.org)
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[filmscanners] Re: film and scanning vs digital photography
You are correct about the "sweat spot" of a lens. This is where the
image field is reasonable flat and distortion is minimal. A few years
ago my company, along with a well respected research lab, conducted
tests on many commercially available lenses. The research lab was
focused on lens systems for video cameras, as was my company, in the
beginning of the study, but we expanded to testing lenses for 16mm mopic
and 35mm still camera lenses that covered a very broad spectrum of
brands. Our tests were to measure the on axis longitudinal chromatic
distortion over most of the visible spectrum - wavelengths ranging from
480 angstroms (blue) to 620 angstroms (red). Using collimated light,
and with the lens collimated to 540 angstroms (D line), we found that
most of the lenses we tested had a displacement of .005" to .007"
between the focus point of blue and the focus point of red. The blue
focus point was short of the D line focus point and the red focus point
was well beyond the D line focus point. In film tests, these lenses
rendered fairly poor resolution at full aperture (about 20 lp/mm) but
improved, to a point, as the lens was stopped down. At some point,
diffraction limit overtakes the chromatic aberrations and the resolution
begins to decline. This is why there is also a sweat spot on the
aperture scale - that aperture where resolution peaks and before
diffraction limit takes over.
To use a lens with .006" to .007" chromatic displacement on todays
digital cameras with their smaller apertures would not produce tack
sharp images. They weren't that sharp on any camera at full aperture.
Lens systems designed for today's digital cameras seem to produce very
sharp images. Admittedly, the lenses I have are not fast lenses (f:4
that ramp up to f:5.6 and max focal length).
By the way, the best lenses that we tested measured a span of .002" to
.004" between blue and red. They included Angenieux, Kern, Leitz, and
Zeiss. Although I haven't tested any of the digital camera lenses, I
would expect that they are closer to these tolerances.
Arthur Entlich wrote:
> I honestly don't know enough about lens/optical mechanics to tell you
> what the limitations are.
> As I understand it, the "sweet spot" of most lenses in terms of
> resolution is toward the middle of the lens, which is the area used on a
> smaller sensor camera using lenses designed for larger frames.
> However, when we get into issues like light falloff and losses due to
> the angle of the light striking the sensor, I have no idea which becomes
> more critical in the end result.
> Hanna, Mark (x9085) wrote:
>> This makes good sense Art, however I'm curious about pixel density.
>> (apart from the obvious larger pixel = more photons landing in it
>> sensitivity advantage which is often the case with the larger sensor)
>> Can the lenses being used on the cameras in question, satisfactorily
>> resolve the number of lines per mm required for the smaller pixel
>> density of the smaller sensor?
>> I have read about lenses having 40LPmm (crap consumer zoom)or 100LPmm
>> (reasonably good lens), is this figure in relation to the intended
>> projected plane? If so, 40LPmm for a 35mm film plane or FF sensor would
>> be 24mm by 36mm which at 40LPmm, equals 1.3824 MPixels. 100LPmm =
>> For an APSC sized sensor, 15 by 24mm I think, you're looking at 0.576MP
>> and 3.6MP for 40LPmm and 100LPmm respectively.
>> So in theory, you may be able to crop the FF pic to emulate a 1.3 or 1.6
>> sized sensor, and despite possibly having less pixel density, the sensor
>> may be capturing the same actual sharpness or resolution, in which case
>> you could simply upsize the resolution to match in PS, and get the same
>> resolution, same sharpness, but lower noise photograph, due to larger
>> pixels, but pixels that may actually match the resolution of the lenses
>> better than the smaller sensor.
>> I don't know much about lens resolution, however if the average L series
>> lens is around 100 to 120LPmm, I know I'd be wanting the larger sensor
>> if my above assumptions are correct. I have a 5D, and the size and
>> resolution of the images never fail to amaze me, as good as my old
>> Mamiya M6451000S.
>> -----Original Message-----
>> From: email@example.com
>> [mailto:firstname.lastname@example.org] On Behalf Of Arthur Entlich
>> Sent: Wednesday, 11 July 2007 9:47 AM
>> To: Hanna, Mark (x9085)
>> Subject: [filmscanners] Re: film and scanning vs digital photography
>> Let's say you have two sensors, each 12 MP. One is FF the other smaller
>> using 1.3X factor. To get the same multiplication factor with the FF,
>> you have crop about 1/4th of the area out, which means you have reduced
>> the resolution by that much. If the FF is about 1/4th higher res to the
>> smaller sensor, then you are correct, no disadvantage.
>> Considering cost and weight of a FF, may not be as great an advantage as
>> it first appears.
>> gary wrote:
>>> I simply see no advantage to have a smaller sensor. I don't see how I
>>> spent pixels. This makes no sense to me.
>>> Nikon has an option on some models where you can toss the outer area of
>>> the sensor to save space on the memory card.
>>> R. Jackson wrote:
>>>> Sure, but you "spend" pixels of your total sensor resolution to get
>>>> On Jul 10, 2007, at 9:37 AM, gary wrote:
>>>>> A cropped sensor really doesn't give you more reach. If you think
>>>>> it, you could just crop a full size image to get more "reach."
>> Unsubscribe by mail to email@example.com, with 'unsubscribe
>> or 'unsubscribe filmscanners_digest' (as appropriate) in the message
>> title or body
>> This email, and any attachments transmitted with it, is confidential and may
>> contain sensitive or privileged information. If you are not the named
>> recipient you may not read, use, copy, disclose, distribute or otherwise act
>> in reliance of the message or any of the information it contains. If you
>> have received the message in error, please inform the sender via email and
>> destroy the message. Opinions expressed in this communication are those of
>> the sender and do not necessarily represent the views or policy of Crown
>> Castle Australia Pty Ltd. No responsibility is taken for any loss or damage
>> sustained from the use of the information in this email and Crown Castle
>> Australia Pty Ltd makes no warranty that this material is unaffected by
>> computer virus, corruption or other defects.
Unsubscribe by mail to firstname.lastname@example.org, with 'unsubscribe
or 'unsubscribe filmscanners_digest' (as appropriate) in the message title or